How to calculate the carbon footprint of a product and document the result
Three people review a metal housing at a workbench with papers and a laptop. Machined pieces and industrial equipment are visible around them.
Diagram text description
The carbon footprint of a product expresses the greenhouse gas emissions attributable to a defined product over a stated life cycle boundary, usually in kilograms of carbon dioxide equivalent, or kg CO₂e, per specified unit. Start with a question the result should answer. An engineer deciding which production step to improve needs a different level of detail from a buyer evaluating two products for the same job.
A product footprint focuses on climate impact. A broader life cycle assessment can also examine impacts such as water use and effects on ecosystems. A company inventory accounts for an organisation’s activities over a reporting period. It can supply relevant energy or purchasing records, but those records need to be attributed to the product before they can support its footprint.
A useful result tells the reader what was counted, what it represents and which assumptions could change it. The example below calculates selected emissions for one packed component and then shows what the evidence file would need before anyone reused the figure. Keep the gross inventory separate from decisions about offsets or public claims.
Start with the product, unit and purpose
Define the product and the decision together. Begin with the SKU, which is the identifier used to order or track the item. Add its configuration, revision and production site. If the same SKU can come from two factories, identify whether the result represents one site or a stated mix of production. Use the period of the data to describe what the result represents.
| Specification | What to record |
|---|---|
| Product identity | SKU, version, material configuration, manufacturing site and packaging format |
| Unit of analysis | For example, one packed component leaving a named factory gate |
| Time represented | Production period and periods of material, energy and freight data |
| Intended decision | Internal hotspot review, supplier data request, or another defined use |
| Performance reference | Function, expected use, required performance and service life where the study depends on them |
Table 1, shown one row at a time. Every field is retained.
Row 1
- Specification
- Product identity
- What to record
- SKU, version, material configuration, manufacturing site and packaging format
Row 2
- Specification
- Unit of analysis
- What to record
- For example, one packed component leaving a named factory gate
Row 3
- Specification
- Time represented
- What to record
- Production period and periods of material, energy and freight data
Row 4
- Specification
- Intended decision
- What to record
- Internal hotspot review, supplier data request, or another defined use
Row 5
- Specification
- Performance reference
- What to record
- Function, expected use, required performance and service life where the study depends on them
For a calculation of selected production stages, one packed component can be a clear declared unit. A comparison of two guards with different lifetimes needs to consider the protection they provide over the same service period, including replacements where relevant. Equal quantities of items need not deliver equivalent performance.
The fictional calculation identifies AG-12 revision C, one packed guard at the factory gate. The analyst must also keep period, site, boundary, evidence quality and the approved use. A mixed product family or a shifting unit needs clarification before a single result can be reused.
Diagram text description
Write the specification at the top of the calculation and the handoff. If the product or purpose changes, decide whether the original unit and evidence still fit before reusing the value.
Choose a life-cycle boundary you can defend
Draw the stages from material production through manufacture, delivery, use and end of life. Mark the point where the model begins and ends. A cradle-to-gate boundary extends from upstream resource extraction and production to the stated gate. A cradle-to-grave boundary extends through use and end-of-life treatment. Where a preliminary estimate covers selected activities only, name those activities explicitly.
| Stage | What to define |
|---|---|
| Materials and packaging | Inputs, upstream production covered by their factors and treatment of losses |
| Inbound transport | Origin, destination, mode and the legs already included in supplier factors |
| Manufacturing and packing | Processes at the site, shared activities and what leaves the gate |
| Distribution and use | Routes, storage and assumptions about installation, maintenance or operating energy |
| End of life | Treatment routes, geography and the method used for waste and recycling |
Table 2, shown one row at a time. Every field is retained.
Row 1
- Stage
- Materials and packaging
- What to define
- Inputs, upstream production covered by their factors and treatment of losses
Row 2
- Stage
- Inbound transport
- What to define
- Origin, destination, mode and the legs already included in supplier factors
Row 3
- Stage
- Manufacturing and packing
- What to define
- Processes at the site, shared activities and what leaves the gate
Row 4
- Stage
- Distribution and use
- What to define
- Routes, storage and assumptions about installation, maintenance or operating energy
Row 5
- Stage
- End of life
- What to define
- Treatment routes, geography and the method used for waste and recycling
Separate a stage excluded by the declared study boundary from an activity omitted because its data are missing. The first defines the scope. The second is a gap that may affect the estimate within that scope. Record both and explain how each limits the result.
Included activities are aluminium and carton production, inbound transport and allocated machine electricity. Facility heat and offcut treatment are omitted from the selected estimate. Outbound distribution, use and end of life are outside it. Describe the selected activities and limitations with the figure.
Diagram text description
For a production improvement project, the team might begin with a gate-based estimate. If the resulting design change affects maintenance, use or replacement, assess those effects before claiming an improvement across the life cycle.
Gather the activity and supplier evidence
Work from the bill of materials, or BOM, for the version being made. Check that the supplier’s part number and grade match the purchased item. Ask production which inputs the BOM omits because they are consumed during manufacture or treated as shop supplies. Add packaging and track production losses so material input and finished output can be reconciled.
| Activity | Evidence to obtain | Review check |
|---|---|---|
| Material input | Versioned BOM, issue record, weighing or yield records | Match material grade and input mass to the product revision |
| Supplier production | Product-specific factor or underlying data, site, period and boundary | Check the unit and whether upstream inputs and transport are covered |
| Process energy | Meter reading, site factor and production or machine logs | Confirm the period and distinguish directly metered from allocated quantities |
| Inbound freight | Shipment mass, route, mode and carrier data | Confirm which transport legs the material factor includes |
| Packaging | Actual packaging specification and mass | Include inserts or other pieces used for the declared packed unit |
Table 3, shown one row at a time. Every field is retained.
Row 1
- Activity
- Material input
- Evidence to obtain
- Versioned BOM, issue record, weighing or yield records
- Review check
- Match material grade and input mass to the product revision
Row 2
- Activity
- Supplier production
- Evidence to obtain
- Product-specific factor or underlying data, site, period and boundary
- Review check
- Check the unit and whether upstream inputs and transport are covered
Row 3
- Activity
- Process energy
- Evidence to obtain
- Meter reading, site factor and production or machine logs
- Review check
- Confirm the period and distinguish directly metered from allocated quantities
Row 4
- Activity
- Inbound freight
- Evidence to obtain
- Shipment mass, route, mode and carrier data
- Review check
- Confirm which transport legs the material factor includes
Row 5
- Activity
- Packaging
- Evidence to obtain
- Actual packaging specification and mass
- Review check
- Include inserts or other pieces used for the declared packed unit
Reconcile receipts with the activity actually used. A catalogue may give finished mass while the calculation needs the quantity fed into manufacture. If 0.90 kg enters the process and 0.80 kg leaves in the finished component, investigate the 0.10 kg difference and document how the waste is treated. Using the finished mass alone can understate upstream material production.
Save the factor source, dataset or supplier reference, version, region and original unit beside the value. Preserve the original supplier response and any conversion in the evidence file. If the factor is older than the production data, record why it remains representative or what evidence should replace it.
Procurement maps the order code to the actual supplier item. Engineering confirms the revision, material and packing. Request the source data for the relevant grade and site. Keep its original unit, period, source and any conversion or unresolved field with the calculation. Assign method ownership and independent review.
Diagram text description
Calculate emissions without hiding allocation choices
Consider this fictional example. Merriford Components makes a batch of 100 AG-12 gauge guards, revision C, during a production run in 2025. The declared unit is one finished guard packed at the factory gate. The illustration includes aluminium and carton production, inbound transport for those inputs and allocated machine electricity. All quantities, factors and source references in the example are invented.
The shared machine used 60 kWh across two runs with similar loads. The guard run occupied one of four machine hours, so the allocation assigns it 15 kWh and the other run 45 kWh. The diagram shows that split. The guard batch used 15 kWh for 100 finished units, or 0.15 kWh per guard. Machine time is the proxy for electricity in this example. If loads differ materially, meter by run or find a better physical basis.
The two fictional runs used 60 kWh over four hours at similar loads. Allocate one quarter, 15 kWh, to the guard run and three quarters, 45 kWh, to the other run. Divide the guard share by 100 for 0.15 kWh per packed guard before applying the electricity factor.
Diagram text description
| Included source | Activity per packed guard | Hypothetical factor | kg CO₂e per packed guard |
|---|---|---|---|
| Aluminium production | 0.90 kg input, including 0.10 kg of offcut | 8.00 kg CO₂e per kg | 7.2000 |
| Carton production | 0.08 kg carton | 0.75 kg CO₂e per kg | 0.0600 |
| Inbound truck transport | 0.00098 tonnes × 200 km = 0.196 tonne-km | 0.10 kg CO₂e per tonne-km | 0.0196 |
| Machine electricity | 0.15 kWh allocated to one guard | 0.40 kg CO₂e per kWh | 0.0600 |
| Total for selected activities | Sum of the four rows | — | 7.3396 |
Table 4, shown one row at a time. Every field is retained.
Row 1
- Included source
- Aluminium production
- Activity per packed guard
- 0.90 kg input, including 0.10 kg of offcut
- Hypothetical factor
- 8.00 kg CO₂e per kg
- kg CO₂e per packed guard
- 7.2000
Row 2
- Included source
- Carton production
- Activity per packed guard
- 0.08 kg carton
- Hypothetical factor
- 0.75 kg CO₂e per kg
- kg CO₂e per packed guard
- 0.0600
Row 3
- Included source
- Inbound truck transport
- Activity per packed guard
- 0.00098 tonnes × 200 km = 0.196 tonne-km
- Hypothetical factor
- 0.10 kg CO₂e per tonne-km
- kg CO₂e per packed guard
- 0.0196
Row 4
- Included source
- Machine electricity
- Activity per packed guard
- 0.15 kWh allocated to one guard
- Hypothetical factor
- 0.40 kg CO₂e per kWh
- kg CO₂e per packed guard
- 0.0600
Row 5
- Included source
- Total for selected activities
- Activity per packed guard
- Sum of the four rows
- Hypothetical factor
- —
- kg CO₂e per packed guard
- 7.3396
The two material factors end at their suppliers’ gates. Both inputs travel the same assumed 200 km route together, so their combined 0.98 kg is counted once in the inbound row. For each row, the factor denominator matches the activity unit. Keep the unrounded sum for review and communicate the result as about 7.34 kg CO₂e per packed guard for the selected activities.
The model leaves facility heat, offcut treatment, outbound distribution, use and end of life outside the estimate. It includes the upstream production of all 0.90 kg of aluminium and makes no recycling credit adjustment. The omissions and hypothetical factors limit it to a worked inventory calculation.
Show data quality and uncertainty beside the number
Put the evidence classification beside each input so the reviewer can see what is known. Activity measurements, supplier data and secondary factors can appear in the same row. A metered quantity can still be combined with a factor that has weak geographic or technological fit.
| Evidence class | Example | What to check |
|---|---|---|
| Measured activity | Weighed input or electricity meter reading | Instrument, scope, period and whether the data represent normal production |
| Supplier-specific information | A factor for the actual alloy and production site | Boundary, allocation, coverage, source version and scope of any assurance |
| Secondary estimate | Regional aluminium dataset or average truck factor | Geographic and technological fit, age, denominator and omitted stages |
| Unresolved input | Facility heat without a product allocation | Likely significance, owner, plan to resolve and effect on permitted use |
Table 5, shown one row at a time. Every field is retained.
Row 1
- Evidence class
- Measured activity
- Example
- Weighed input or electricity meter reading
- What to check
- Instrument, scope, period and whether the data represent normal production
Row 2
- Evidence class
- Supplier-specific information
- Example
- A factor for the actual alloy and production site
- What to check
- Boundary, allocation, coverage, source version and scope of any assurance
Row 3
- Evidence class
- Secondary estimate
- Example
- Regional aluminium dataset or average truck factor
- What to check
- Geographic and technological fit, age, denominator and omitted stages
Row 4
- Evidence class
- Unresolved input
- Example
- Facility heat without a product allocation
- What to check
- Likely significance, owner, plan to resolve and effect on permitted use
Test a consequential assumption. In the fictional calculation, a change of 1.00 kg CO₂e per kg in the aluminium factor would move the result by 0.90 kg CO₂e per guard, with the other inputs held fixed. That shows why the alloy route and its factor matter. This sensitivity check describes the effect of one assumption. It does not establish a statistical confidence interval or a verified range.
With 0.90 kg of aluminium input, a change of 1.00 kg CO₂e per kg in the factor changes the selected result by 0.90 kg CO₂e per guard while other inputs stay fixed. It is a one-assumption sensitivity check. Check the actual alloy route, production site and period before reuse.
Diagram text description
If better evidence could reverse the decision, keep the result provisional for that use. Name the assumption to resolve and who can obtain the supporting record.
Use the result to make a useful decision
Use a footprint first to ask where a change could matter. In the guard example, aluminium production dominates the selected estimate, so the buyer could request the actual alloy production route and the design team could examine yield. A production change needs records of material input and equivalent finished output. Compare those quantities on the same basis before attributing a lower result to the design.
| Before comparing options | Check both results for |
|---|---|
| Equivalent function | Required performance, expected lifetime and replacements over the same service |
| Matching reference unit | The same quantity of product or service, including packing if relevant |
| Comparable system boundary | Equivalent included stages and a treatment of material exclusions |
| Consistent methods | Compatible allocation, emissions factors, use assumptions and applicable product rules |
| Reviewable evidence | Comparable data periods, representativeness and an appropriate level of review |
Table 6, shown one row at a time. Every field is retained.
Row 1
- Before comparing options
- Equivalent function
- Check both results for
- Required performance, expected lifetime and replacements over the same service
Row 2
- Before comparing options
- Matching reference unit
- Check both results for
- The same quantity of product or service, including packing if relevant
Row 3
- Before comparing options
- Comparable system boundary
- Check both results for
- Equivalent included stages and a treatment of material exclusions
Row 4
- Before comparing options
- Consistent methods
- Check both results for
- Compatible allocation, emissions factors, use assumptions and applicable product rules
Row 5
- Before comparing options
- Reviewable evidence
- Check both results for
- Comparable data periods, representativeness and an appropriate level of review
The GHG Protocol’s Product Standard overview distinguishes hotspot work from product comparisons that need more specific rules. If one supplier reports material and manufacture while another includes use and disposal, a lower number alone cannot support a purchasing comparison. Ask for a stage breakdown and compatible assumptions before drawing a conclusion.
Match performance, lifetime and any replacements over the same service. Use an aligned unit and set of included stages, consistent methods and reviewable data. If one result includes use and disposal while another ends at manufacture, request a stage breakdown before comparing.
Diagram text description
For internal tracking of the same item, record which change came from the process and which came from an updated factor or allocation method. If the baseline can be recalculated consistently, keep both versions and compare the aligned results.
A practical product evidence handoff
Consider this fictional example. The analyst gives procurement and the internal reviewer the following handoff for Merriford’s guard. The record is marked provisional for internal screening. Each version retains its evidence references and review state.
| Handoff field | Filled entry |
|---|---|
| Product and unit | AG-12 revision C, one packed guard at the manufacturing gate |
| Activity period and file | 2025 batch MC-117, calculation revision 1 |
| Result and included activities | About 7.34 kg CO₂e per packed guard for aluminium, carton, inbound transport and machine electricity |
| Evidence references | BOM C, material issue log M-117, shared meter sheet E-117, run log H-117 and assumed factor sheet F-3 |
| Assumptions and gaps | Alloy factor is a placeholder. The machine share is allocated by hours. Freight distance and factor are assumed |
| Exclusions | Facility heat, offcut treatment, outbound distribution, use and end of life |
| Owners and next evidence | Procurement requests the alloy and plant factor. Production checks the load assumption. The inventory reviewer decides whether to revise or release |
| Permitted use and refresh | Internal hotspot discussion. Reopen when the alloy source, BOM, site or method changes |
Table 7, shown one row at a time. Every field is retained.
Row 1
- Handoff field
- Product and unit
- Filled entry
- AG-12 revision C, one packed guard at the manufacturing gate
Row 2
- Handoff field
- Activity period and file
- Filled entry
- 2025 batch MC-117, calculation revision 1
Row 3
- Handoff field
- Result and included activities
- Filled entry
- About 7.34 kg CO₂e per packed guard for aluminium, carton, inbound transport and machine electricity
Row 4
- Handoff field
- Evidence references
- Filled entry
- BOM C, material issue log M-117, shared meter sheet E-117, run log H-117 and assumed factor sheet F-3
Row 5
- Handoff field
- Assumptions and gaps
- Filled entry
- Alloy factor is a placeholder. The machine share is allocated by hours. Freight distance and factor are assumed
Row 6
- Handoff field
- Exclusions
- Filled entry
- Facility heat, offcut treatment, outbound distribution, use and end of life
Row 7
- Handoff field
- Owners and next evidence
- Filled entry
- Procurement requests the alloy and plant factor. Production checks the load assumption. The inventory reviewer decides whether to revise or release
Row 8
- Handoff field
- Permitted use and refresh
- Filled entry
- Internal hotspot discussion. Reopen when the alloy source, BOM, site or method changes
A buyer could use this handoff to ask for better alloy evidence or to discuss changes in yield. A public footprint or a claim that the guard outperforms a rival remains unsupported. Keep the previous result visible when a supplier sends a new factor and record which rows and limitations changed. Give the reviewer enough information to rebuild the total without asking the original analyst to interpret the spreadsheet.
Where software helps and where people decide
A tool can reduce repeat preparation when it keeps product identifiers, BOM versions and source references together. Before choosing a tool, run one product through it with a unit conversion, a shared input and a missing supplier value. Check whether the reviewer can inspect each original value and see why it was accepted or held.
| Record or decision | What software can support | What the owner must decide |
|---|---|---|
| Product mapping | Match catalogue identifiers and track revision history | Which configuration and site the study represents |
| Factors and conversions | Store units, source versions and calculation steps | Whether the factor and denominator fit the specific activity |
| Shared inputs and omissions | Record allocations, exclusions and change history | Whether the allocation and boundary are defensible |
| Review and release | Keep review status, comments and exports | What uses or claims the evidence supports |
Table 8, shown one row at a time. Every field is retained.
Row 1
- Record or decision
- Product mapping
- What software can support
- Match catalogue identifiers and track revision history
- What the owner must decide
- Which configuration and site the study represents
Row 2
- Record or decision
- Factors and conversions
- What software can support
- Store units, source versions and calculation steps
- What the owner must decide
- Whether the factor and denominator fit the specific activity
Row 3
- Record or decision
- Shared inputs and omissions
- What software can support
- Record allocations, exclusions and change history
- What the owner must decide
- Whether the allocation and boundary are defensible
Row 4
- Record or decision
- Review and release
- What software can support
- Keep review status, comments and exports
- What the owner must decide
- What uses or claims the evidence supports
Test how a correction works. If procurement replaces a factor, can a reviewer recover the original file, see which products were affected and explain what changed in the reported result? If the new factor applies to only one site or grade, verify that it stays tied to that product record. A convenient import cannot establish that the supplier’s boundary or method matches yours.
Give unresolved values a status and an owner so automation does not make them appear final.
Prepare one product record for review
Begin the handoff with one catalogue line and its current manufacturing record. Procurement maps the order code to the supplier’s item. Engineering checks the revision and BOM against what is built. The analyst lists the activity and factors already available, then identifies which evidence must come from the supplier or the production team.
For example, if a purchasing code covers both a bare guard and a packed kit, split the mapping before calculating. A pack list can show which carton and inserts belong to each configuration. If the component itself changes alloy or site under the same code, keep the variants identifiable or explain how the production mix is represented. A revision field is useful only when the affected calculation refers to it.
Assign a method owner and a reviewer before the handoff leaves the team. The method owner documents the unit, boundary and calculation choices. The reviewer checks whether the records support the intended use and sends unresolved inputs to the person who can obtain them. Start with the evidence request most likely to change the decision. In the guard example, that is the factor for the actual alloy and plant, with its boundary and period.
What to check before sharing a product footprint
Read the result and its supporting file together before sharing it. Choose a level of review appropriate to the intended use, and record which use the reviewer approved.
| Release check | Hold the figure for that use when |
|---|---|
| Product, period and unit are explicit | The SKU combines different products or the unit is ambiguous |
| Included stages and exclusions are traceable | A relevant stage or overlap remains unresolved |
| Values and allocation can be reproduced | A key factor, conversion or shared-process method lacks evidence |
| Uncertainty and review status are visible | A material assumption could change the claim or comparison |
| Wording matches the approved use | An internal screening estimate would be presented as a public product footprint |
Table 9, shown one row at a time. Every field is retained.
Row 1
- Release check
- Product, period and unit are explicit
- Hold the figure for that use when
- The SKU combines different products or the unit is ambiguous
Row 2
- Release check
- Included stages and exclusions are traceable
- Hold the figure for that use when
- A relevant stage or overlap remains unresolved
Row 3
- Release check
- Values and allocation can be reproduced
- Hold the figure for that use when
- A key factor, conversion or shared-process method lacks evidence
Row 4
- Release check
- Uncertainty and review status are visible
- Hold the figure for that use when
- A material assumption could change the claim or comparison
Row 5
- Release check
- Wording matches the approved use
- Hold the figure for that use when
- An internal screening estimate would be presented as a public product footprint
For AG-12 revision C retain the 2025 batch, calculation version, declared unit and boundary, BOM and activity, factor assumptions, exclusions, owner and refresh trigger. Software tracks mapping and versions; the reviewer judges scope, factor fit and permitted use. Request the actual alloy factor before considering wider use, and keep offsets separate.
Diagram text description
Send the next evidence request
Keep offset transactions separate from the gross inventory and from the company’s reporting record. For Merriford, the immediate step is to request the matching alloy factor and repeat the review with the documented exclusions in view. Send the figure with the product unit and boundary if it is ready for its intended use. Otherwise share the outstanding evidence request and the date for a new decision.
Company carbon accounting
